Audio Research LS 25 MkII compared to actual AR production ?

Johnny is right, as usual.  Just about every ARC line stage and preamp for which I've seen specs, including the LS25 MkII and the other ARC models that have been mentioned, has a **minimum** load recommendation of 20K, and the input impedance of your Spectral amp is 10K.  Although in both cases it isn't made clear if those numbers apply to the balanced or unbalanced interfaces or both.

In the case of a tube preamp a significant consequence of that sort of mismatch is likely to be perceptible rolloff of the bottom one or two octaves.  That will occur because of the output impedance rise at deep bass frequencies that will be caused by the coupling capacitor that is used at the output of most tube preamps.

Good luck.  Regards,
-- Al
 

In the facts, i.e. listening, LS25mkII sounds great with no roll off or problems.

Good!

The general thumb rule of 10x is respected and the most restricted Audio Research thumb rule of 30x is in the limits.

To clarify, though, to assure impedance compatibility the 10x rule of thumb guideline should be applied at the frequency for which the output impedance of the component providing the signal is highest. Most impedances are specified at a mid-range frequency such as 1 kHz. It is very common for tube preamps to have output impedances at deep bass frequencies that are much higher than that specified value, often 2K or 3K or even 4K ohms. That rise at low frequencies results from the output coupling capacitor that is used in the majority of tube preamps (and also in some solid state preamps), as I mentioned in my previous post. The impedance of a capacitor increases as frequency decreases.

If as is often the case the component’s output impedance at 20 Hz is not known, and is not indicated in published measurements (such as Stereophile often provides), then to be safe a considerably higher ratio than 10x should be used, something like 50x or 75x IMO. Especially if the component is tube-based and is likely to have a coupling capacitor at its output.

Also, to clarify a common misconception I should add that failing to meet that guideline does not necessarily mean that there will be an impedance compatibility problem. It depends on how much **variation** there is in the output impedance over the frequency range. But meeting that guideline (at all audible frequencies) assures that there won’t be an impedance compatibility problem.

In this specific case, I took a look at the schematic for the LS25 MkII at arcdb.ws, and based on some quick calculations it appears that its unbalanced output impedance at 20 Hz is in the vicinity of 1.3K or so. A bit more than ideal relative to a load impedance of 10K, but reasonably consistent with your finding that it is not an issue.

Regards,
-- Al